Jgro finished

[dart70ca]

Buddy and i finished our JGro machine this morning. Want to post some pics but I don't see anywhere to attach them here. I expect one of you will help me out !
We bought the 4-axis board, power-supply and fan kit from Xylotex. 269 oz. in. motors from AutomationDirect (NEMA 23's). Leadscrews were ¼ - 20 threaded rod. We used flange-bearings (wheel-barrow wheel bearings) on each end of the x-axis screw with nuts and washers to tension it and reduce whip. These bearings are poor choices and do NOT take the axial thrust required here. Will replace later. Bearing-block bearings came from VXB at the most excellent price of $21 US for 30 pcs delivered to my door within 5 days of ordering (and there was a weekend and a border-crossing in there!).
Design and drawings were pretty good. Only things we changed were the addition of plastic adjuster blocks for the big pipes of the x axis; the MDF ones stripped their threads during setup. We also changed the leadscrew tensioning setup for the x-axis to take all tension loading off the motor as steppers don't usually have bearings to take thrust loads like that. Easy enough to do. Used some cheap Lovejoy-type jaw-couplings from www.princessauto.com to join motors to leadscrews. Spaced the motors away from the machine body with a couple pieces of carefully cut 2" PVC pipe and long mounting bolts. 10-24 nuts will NOT fit very well at the motor housing so we would suggest long screws here rather than threaded rod which we were forced to use.
So far we have only used a Dremel tool on it but it worked well. Looking forward to a router. We get about 30" a minute out of it which is about all I'd want to push a Dremel at. Lot of guys said we wouldn't be happy with the poor performance of the NEMA 23's but we think they're just fine. Not trying to win any races here.
Some cautions we have for others building it are as follows;
1) Check your stock size. The drawings are based on .75 thick board. Ours was somewhat less than that at .72 and this caused some alignment and fit problems between the bearing blocks and pipe guides, among other things.

2) Didn't have time to check if it was our mistake or not yet, but the x and y axis lead screw nuts seem to be mounted a bit higher than the motors and outboard leadscrew supports. This causes binding and motor overloads near the limits of travel n those axes. Check the drawings before cutting the mounting holes.

3) While tapping the MDF made us nervous and skeptical at first, it actually works pretty well. It works quite a bit better in the face of the stock than through the edges. Be prepared to drill a clearance hole and put a nut on the other side or make a new part; these DO strip out quite easily. If you just assemble and leave it, it should be OK. Repeated disassembly and assembly would not be recommended here. Use a good drill bit and a fresh sharp tap of good quality.

4) We spent a day and a half diagnosing a weird motor reversing symptom and traced it back to how we wrapped the wires. The motors are bipolar units and the wires for each phase should be kept separate. We twisted all 4 wires from the motor together along their whole length. This caused (we assume) an induced charge in adjacent pairs of wires and this in turn caused the motors to change direction after 5 movements or less. eg; we jogged for 1 second intervals, 10 times and the motor randomly reverses direction 3 times and would often lock up before 15 jogs. We separated the four wires into 2 twisted pairs and took the extra precaution of wrapping them in foil. Problem solved.

5) We attempted to use EMC to drive this machine (initially) because, 1) we have a good friend who is Linux-fluent and offered to det it up for us, 2) it's free and 3) supposed to be pretty stable once set up. We had this old donor-computer (333 PII MMX) that we thought would do. We wanted 3 partitions on the HD so we could mount some version of windows and DOS as well so we could ise Mach2 or TurboCNC if necessary. To make a long story short, Linux and EMC got in, Windows and DOS didn't. We ended up installing WinXP in a single partition and running Mach2. We had a lot of trouble finding a version of DOS for download to try TurboCNC. In any event, we wanted to use a thumb-drive or wireless LAN to transfer programs from house computer to shop and neither DOS nor Linux (?) have drivers for these (yet). Least not ones that could be installed and tweaked by us Neanderthal's. House computer doesn't have a floppy and we don't want one. It's still to be seen whether or not Mach2 will run consistantly on this old machine but so far so good.

That's about all for now. There's a bit more I should get down while it's fresh but I'll save it for when i put the photo's up.

Many thanks to Jgro for doing up these plans for us and to Mr. Keinbauer for the inspiration (and cajoling).

Keith Green, Surrey, BC
Ray Webster, Kamloops, BC
and
Rob McCrea, Langley, BC

[abasir]

Use the "Manage Attachments" to attach your photos...

[dart70ca]

Thanks. Here thaey are.

Keith

[whirlybomber]

Nice looking machine.

I'm part way through glueing up the base for my Jgro a.t.m.

Some newb questions here,

1) how do you make sure the axis are perpendicular to each other? By eye doesn't sound good to me, and I don't know if my skill with a square is up to speed. I suppose I coud borrow machined square blocks of someone.....

2) how did you level the cutting bed? I was looking for something a little more elegant than running the router over it.

Cheers,
Brad

[Jason Marsha]

Good job guys.

To solve the binding problem at the limit of the axis travel you may need mount the bearings outside of the gantry wall and make the necessary adjustments.

I had a similar problem and it became very obvious when I changed to 5/16-18 threaded rod which had a lot less give than the 1/4-20.


Jason

[Weldtutor]

Keith, Ray, Rob
I like the cable carrier you used! Great machine. Thanks for the posts & photos. Princess Auto is one of my favourite haunts.
WT

[dart70ca]

Preliminary alignments were done with a digital vernier measuring from the edge of the assembly to the nearest point on a pipe. Our components were cut on my buddie's Unisaw and that thing cuts really straight and square so I didn't have much trouble doing it this way. Haven't run many parts yet to see how aligned the table actually is.
What I meant to do but didn't have time before I had to leave (we built this at my buddie's shop which is 350 km from here) was drill a hole on the Dremel/router mount the appropriate size for a rod from my dial indiactor base. Tap a setscrew into it to hold the rod and you can run a dial indicator up and down to find the high and low spots before running the router over it. I suppose you could do it before you secured the table and shim it if you wanted.
You should get away from thinking that the simple solution is the inelegant one. Running the router over it is the quickest, simplest way to flatten your table. That's the way we do it in industry, too. No matter how hard you try to square it up, you will always be working with a $10 pice of compressed sawdust. Don't waste your time.

I want to add a little about wiring the motors again here. I just got a reply from a guy on the Xylotex group says he uses a microphone cable with the trade-name of StarQuad. It has four heavily shielded conductors in guages up to 18. He's getting away with 24 or 26 ga. Hard to believe, but...
I like the microphone cable idea because 1) it must be shielded well to work in the environments it was designed for, 2) it has to be very flexible and remain so, as you can imagine the abuse such a cable would take. 3) You can get them pre-assembled with sturdy plugs and the mates for them can be had in a panel-mount configuration if you want.

Keith Green
Surrey, BC

[csmarmot]

Keith, et. al.,

I'm just about ready to take the plunge into starting my project, and I have to say each new post like yours psyches me up a bit more. Just about over the edge. Thanks for passing on the microphone cable tip. Looking forward to hearing about your progress and maybe seeing some cuts.

cheers,
Chris

[dart70ca]

What you've done there is basically what we did. I even contemplated using the coupling nuts as standoffs for the motor! What bearings have you used for the ends there? Those wheelbarrow flange-bearings we used suck pretty bad (they were $1 each). How much up or down did you move the mounts to get them in alignment?

Keith

[dart70ca]

The carrier was made by Ray (woodworker and owner of the Unisaw). He just ripped a few scraps of MDF so they were real thin and built a long trough for the cable. Covered the top for half it's length with another strip. If you rip that stuff real thin it has quite a bit of flex and you can even make one of those fancy curved cable-carriers from it if you want. Our cables were too short to bother with that. The wires themselves are wrapped in that corrugated split black plastic crap from Canadian Tire. You will need 2 packages 10ft each to do this machine up. We used the ¾" stuff. Below the cable tray there is a smaller box he made up after I left. He discarded the metal box I bought for the controller card and put it in there. Might be more susceptible to static and dirt, though. Haven't seen it close up yet. He did it so that all we have to do to remove the computer from the system is to disconnect the printer cable. When the shop goes into cabinet-building mode it's too dusty and crowded to keep the CPU and monitor in there. By the way, we picked up that monitor, keyboard and optical mouse for $20 for the lot from a local computer store. New Years Blowout sale. At those prices (and they're out there) you can afford to be hard on the things a bit.

Keith

[dart70ca]

I have to admit, I've been waiting a long time to post those pic's here. It's stuff like that that got me going, too. I'll get some closeups of details and maybe some cuts up asap.
The machine was built mostly for Ray to use. He's a G-scale model railroader and wants to use it to help build custom cars and buildings. We did a scale drawing of a 40 x 40 ft. brick wall as a test to see what we could use the machine for. Comes out at about 17" square (29:1 scale) and contains about 4000 bricks. .017 wide grout lines by about .01 deep. You don't need a very strong or fast machine to make something like that, but I would say that a CNC is certainly the way to go and you need something that will do at least 2 feet square.

Keith

[dart70ca]

Now that the machine is finished more or less, what's a good (read that as 'inexpensive') source for carbide Dremel bits? I found the www.microcutusa.com website this morning and it looks really good. Anyone else have a link to pass on?

Keith